Engram

MCP serverEverything else

Shared memory for Claude, ChatGPT and any MCP client. Read and edit it in Obsidian or the web app.

Available today. Use it from your connected AI after setup.

Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.

Then ask your AI: use Engram

Install Engram

The server’s own address, for the clients that take one directly. Or connect ahel onceand every client you use reads it from one address, with the account kept on ahel rather than in each client’s config.

  • Claude Code

    claude mcp add --transport http --scope user engram 'https://mcp.engram.page'

    Run it once in your project, then open /mcp to approve any sign-in the server asks for.

  • Claude Desktop

    https://mcp.engram.page

    Add a custom connector in Settings, paste this address, and approve the sign-in.

  • Cursor

    cursor://anysphere.cursor-deeplink/mcp/install?name=engram&config=eyJ1cmwiOiJodHRwczovL21jcC5lbmdyYW0ucGFnZSJ9

    Open the link and Cursor adds the server at that address.

  • ChatGPT

    https://mcp.engram.page

    In Settings, enable Developer mode, create an MCP app, and paste this address. Your plan and workspace must allow custom apps.

  • Codex

    codex mcp add engram --url 'https://mcp.engram.page'

    Run it once, then sign in with codex mcp login engram if the server asks for an account.

From the project's README

As published by engram-app/engram in README.md.

MCP-native · semantic + keyword search · real-time sync · self-hostable · source-available · Elixir/Phoenix

Quickstart · Self-host · MCP · API · Architecture · Plugin · Discord

Your notes stay plain markdown that you and your AI assistants both read and write. Pairs with the Engram Obsidian Sync plugin to sync your whole vault.

Self-Host (Docker Compose)

Don't want to run the commands yourself? Open the Quickstart docs, select the whole page, and paste it into your AI assistant — it's written top-to-bottom as imperatives the AI can execute in order.

Otherwise:

git clone https://github.com/engram-app/Engram.git
cd Engram
cp .env.example .env       # then fill in the three secrets at the top
docker compose up -d

App at http://localhost:4000. Migrations run on boot. Only port 4000 is host-exposed; everything else stays on the private Docker network.

Large vaults? Enable MinIO for S3-style attachments: docker compose --profile s3 up -d — see storage docs.

Better embeddings? Switch to Voyage AI in .env — see embeddings docs.

Full self-host documentation

TopicLink
Quickstarthttps://engram.page/docs/self-host/quickstart/
Environment varshttps://engram.page/docs/self-host/environment-variables/
Encryption & keyshttps://engram.page/docs/self-host/encryption/
Backup & restorehttps://engram.page/docs/self-host/backup-restore/
Upgradeshttps://engram.page/docs/self-host/upgrade/
Troubleshootinghttps://engram.page/docs/self-host/troubleshooting/
Architecturehttps://engram.page/docs/self-host/architecture/
MCP setuphttps://engram.page/docs/mcp/
HTTP APIhttps://engram.page/docs/api/

Community

Questions, ideas, or bug reports? Join us on Discord.

Contributing

Local dev setup, tests, and PR rules: see CONTRIBUTING.md.

License

Dual-licensed:

External contributions sign the Engram CLA. See CONTRIBUTING.md.

Security

See SECURITY.md for vulnerability disclosure. Self-host LAN deployments are out of scope of our published SLA — security depends on the operator's network and infra.

Copyright (c) 2026 Rasbandit Software Solutions LLC d/b/a Engram.

Advanced
Delivery
engram MCP server → your ahel connector (mcp.ahel.ai) → your AI.
Catalog kind
mcp-server
Key
io-github-engram-app-engram
Source
github.com/engram-app/engram
Hosted endpoint
https://mcp.engram.page